Your operations team knows SNMP, YANG, and MIB structures. The next generation of OSS engineering needs streaming graphs, AI correlation, and geospatial telemetry. The hiring profile has changed completely — and almost nobody is hiring for it yet.
By Shawn Ennis • August 18, 2026 • 7 min read
I’m going to make a prediction. The single biggest implementation risk to slice-aware operations in the next five years isn’t going to be technical. The platforms exist. The architecture is proven. The standards are mature. The blocker is going to be people.
Specifically, the people who run the operations stack today have the wrong skills for the operations stack tomorrow. And almost nobody in the industry — vendors, operators, training organizations — is taking the talent transition seriously.
This post is for the VPs of Network Operations, Heads of NOC, and HR partners in telecom who are about to discover that the people they hired five years ago for one job are about to be asked to do a fundamentally different one. The good news is the transition is solvable. The bad news is that nobody has started solving it yet.
The skills that built telecom operations
For 25 years, telecom operations engineering has been a remarkably stable discipline. The core competencies are clear and they don’t change much from year to year:
- SNMP — Simple Network Management Protocol, the foundation of device monitoring
- MIBs — Management Information Bases, the structured definitions of what devices report
- YANG — the modeling language used in modern device configuration
- NetConf — the protocol that exchanges configuration between systems and devices
- Trap handling — knowing how to interpret alert streams
- Topology discovery — understanding how the network connects
- Alarm correlation — identifying which alerts relate to which root causes
- Service modeling — defining what a service IS in a structured catalog
The engineer who knows these things deeply has been the foundation of every NOC and every OSS deployment for a generation. They know the protocols, the data structures, the integration patterns, and the failure modes. They’ve been called in at 2 AM to debug correlation engines that produced the wrong root cause. They’re the institutional memory of how the network actually works.
This skill set is not going away. Static networks still exist. Permanent services still need to be operated. The traditional OSS engineer is going to be valuable for a long time.
But they are no longer sufficient.
The skills that operate ephemeral services
Slice-aware operations require an overlapping but materially different set of competencies. Some of them are extensions of traditional OSS skills. Some are entirely new.
- Streaming data engineering — the ability to reason about telemetry as continuous flows rather than periodic polls
- Graph databases and graph algorithms — because dynamic topology is fundamentally a graph problem
- AI/ML for correlation — understanding how correlation engines learn rather than how they’re configured
- Geospatial data processing — because services move and the OSS must follow
- API-first thinking — every operation must be triggered, monitored, and reported via API
- Event-driven architecture — the operations layer must react to state changes in real time
- Time-series analysis — because ephemeral services produce signals that must be analyzed against their specific time window, not against a long historical baseline
- Service-context modeling — defining what an ephemeral service IS, dynamically, in a way that the correlation engine can use
The engineer who knows these things deeply is rare. They tend to come from one of three backgrounds: data engineering at a software company, observability platform engineering at a large internet operator, or AI/ML infrastructure work. They rarely come from traditional telecom.
The cultural gap that makes this harder
The skills gap is one problem. The cultural gap is harder.
Traditional NOCs run on procedure. Alert fires, follow the runbook, escalate per the matrix, document the resolution, update the runbook. The premium is on discipline, repeatability, and conservative judgment under pressure. These are valuable traits.
Slice-aware operations run on engineering. Correlation engines need tuning, not configuration. Service models need to be designed, not deployed. Anomaly detection needs to learn from data, not from runbooks. The premium is on hypothesis-driven thinking, comfort with statistical reasoning, and the kind of engineering judgment that says “I don’t know what’s wrong but I can build a test to find out.”
These two cultures collide when you try to put a slice-aware operations layer on top of a traditional NOC. The NOC team sees the new platform as fragile, hard to predict, and operationally risky. The engineers running the new platform see the NOC team as too slow to respond to dynamic events. Both views are right. Neither is wrong. But the integration is hard.
Three approaches to the talent transition
Operators are starting to deal with this — quietly, mostly without a coherent strategy. The approaches I’m seeing fall into three categories.
Approach 1: Hire for the new skills externally
Recruit engineers from outside telecom — from cloud infrastructure, observability platforms, or AI engineering. The candidates are out there but they’re expensive and they’re being recruited by everyone. They also have to learn telecom from scratch, which is a serious investment in their first 12-18 months.
Risk: Two-culture problem. The new hires don’t trust the legacy team and vice versa. Strong technical leadership is required to bridge the cultures.
Approach 2: Retrain the existing team
Send your existing OSS engineers through training in streaming data, graph databases, AI correlation, and geospatial analysis. Internal experience compounds beautifully — the people who already know your network can pick up new techniques faster than outside hires can pick up your network.
Risk: Time. Real retraining takes 12-18 months at minimum, and the operator has to commit to the long view. Some engineers won’t make the transition; you need a strategy for the people who don’t.
Approach 3: Build a parallel team
Stand up a dedicated slice-aware operations team alongside the traditional NOC. The two teams report to a common operations leader. Over time, the slice-aware team grows and the traditional team’s scope narrows. Eventually they merge or the traditional team is absorbed.
Risk: Organizational complexity. Two ops teams running in parallel is expensive and creates friction at handoff points. Requires very clear ownership boundaries.
My recommendation: most operators end up doing a combination of all three. External hires for senior technical leadership and architecture. Retraining for the deep telecom experts who can make the transition. Parallel team structure during the transition period, with a planned consolidation 18-24 months in.
The hiring profile worth writing now
If you’re hiring for the new skills, the job description doesn’t look like a typical NOC engineer requisition. It looks more like this:
“Senior operations engineer with experience in streaming data platforms, graph-based correlation, and real-time observability. Background in telecom is a plus but not required. Background in cloud infrastructure operations, observability platform engineering, or large-scale data engineering will receive equal consideration. The role involves operating and tuning AI-augmented correlation engines, modeling ephemeral service contexts, and integrating real-time telemetry across network, edge, and application layers.”
That description is not going to be filled from your existing recruiting pipeline. It’s going to take a different sourcing strategy, a different compensation framework, and probably a different interview process. Start now, because everyone else is going to start in the next 12 months.
The deeper question: The skills crisis isn’t really about hiring. It’s about whether your operations organization is structurally prepared for the transition from static services to ephemeral ones. The technology shift forces an organizational shift. Operators that handle the people change capture the slicing revenue. Operators that don’t keep losing it to platforms run by people who already think this way.
The good news
This transition has happened before. The shift from circuit-switched to packet-switched networks in the 1990s required exactly this kind of skills retooling. The shift from physical to virtualized infrastructure in the 2010s required it again. Each time, the industry adapted. The engineers who navigated the transition successfully became the senior leaders of the next era. The operators that invested early in the talent transition led the cycle.
Slice-aware operations is the third major skills inflection point in the modern telecom era. The pattern is the same. The investments that work today are the same ones that worked last time: hire externally for senior leadership and architecture, retrain internally for depth, run parallel teams during the transition, plan a consolidation 18-24 months in.
The operators that got the first two transitions right are still leading. The ones who waited too long are still catching up. There’s a reason to think this transition will sort the industry the same way.
The full operational blueprint
Including the talent and training implications of slice-aware operations.
About the author Shawn Ennis is the Founder & CEO of Rapax and Citus Technologies. With 25+ years in telecom operations, Shawn previously founded Assure1 (acquired by Oracle in 2021), holds 12 patents in telecom OSS/BSS, and hosts the Transformation Leaders Podcast. Connect on LinkedIn.